Pilot Purgatory¶

Relocate the diagnosis of an initiative that keeps running successful pilots without ever scaling from 'weak evidence' to 'a missing transition mechanism' — the loop is sustained by each pilot's local success, and the fault sits one level up in the decision infrastructure.

Core Idea¶

Pilot purgatory is the organizational pathology in which an initiative repeatedly runs pilots, demonstrations, or proofs-of-concept of the same intervention without an institutional path that converts a successful pilot into adoption at scale. Each iteration produces favourable findings within its own scope; those findings do not propagate into a budget commitment, an operating-model change, an enterprise rollout, a policy update, or a procurement decision. The structural commitment is the absence of a transition mechanism linking demonstration to institutionalization: the pilot subsystem is operationally complete — it has a funding vehicle, a small team, a steering committee, an evaluation write-up — but the escalation handoff to budget ownership, operational integration, or procurement does not exist or is structurally non-functional. The system is locally stable at the pilot state. What sustains the loop is the local success of each pilot: every cycle ends with a favourable write-up that justifies another pilot rather than triggering a scale decision. The pattern is thus preserved precisely because pilots are doing their job; the failure sits one level up, in the decision infrastructure that should convert success evidence into an adoption commitment. This mislocates the diagnosis: practitioners and funders mistake the pattern for insufficient evidence quality — which suggests running better pilots — when the actual cause is missing transition infrastructure, which requires different interventions entirely: named adoption owners with budget authority, stage-gates designed to force a scale-or-kill decision rather than permit a third pilot, procurement vehicles capable of enterprise rollout, and operating-model changes at the receiving organization. McKinsey's 2017 and subsequent reporting on Industry 4.0 manufacturing coined the term for this pattern specifically: industrial companies running dozens of digital-transformation pilots over years without enterprise rollout, producing strong pilot-site results while the transition to production-at-scale repeatedly stalled. The same structure appears in government digital-services programs, health-tech adoption, education-reform programmes, and research-into-practice translation — wherever the organization that runs pilots is structurally separated from the organization that must absorb and operate the scaled capability.

Structural Signature¶

Sig role-phrases:

the demonstration loop — a pilot, proof-of-concept, or evaluation that runs to completion under its own scope, operationally complete (funding vehicle, team, steering committee, write-up)
the absorber system — the organization that must budget, procure, operate, and integrate the scaled capability, structurally separated from the system that runs the pilot
the successful local evaluation — each cycle's favourable write-up, which justifies another pilot rather than triggering a scale decision
the missing transition mechanism — the absent or non-functional escalation handoff (budget ownership, procurement vehicle, governance, named owner) that should convert demonstration into institutionalization
the cycle stability at the pilot state — the loop locally stable, each iteration ending at the same place with no escalation pressure built in, so it does not self-terminate
the diagnostic-misattribution risk — the favourable results tempting a "run a better pilot" reading (evidence problem) when the cause is missing transition infrastructure
the counter-mechanism — a scale-or-kill stage-gate, an adoption owner holding the operating budget, and enterprise-sized procurement that force resolution rather than another pilot

What It Is Not¶

Not an evidence problem. The pilots succeed — each ends with a favourable write-up — so the fault is not weak evidence and "run a better, more rigorous pilot" only reproduces the loop. The binding constraint sits one level up, in the missing transition mechanism that should convert demonstrated success into a budget commitment, a procurement vehicle, a named owner. The favourability of the results is the tell that the gap is structural, not evidential.
Not sunk-cost persistence. Sunk cost is a bias toward continuing because of prior spend; pilot purgatory is the structural absence of the path to commitment, even when commitment would be rational. Debiasing does nothing here, because no one is irrationally clinging to a doomed program — the route to scale simply does not exist to be taken.
Not the valley of death. The valley of death is a financing gap specifically, between R&D and commercialization. Pilot purgatory is transition-mechanism-shaped more broadly: funding may be fully present while the handoff — procurement, governance, operating-model integration, budget ownership — is still missing. Financing is at most one of several links that can be absent.
Not analysis paralysis. Analysis paralysis never decides; pilot purgatory decides favourably — just at the wrong level. Each cycle reaches a clean positive verdict on the pilot and acts on it (by commissioning another pilot), so the failure is not indecision but a decision that loops at the demonstration level instead of escalating.
Not pivot thrashing. Pivot thrashing is repeatedly changing direction; pilot purgatory is repeatedly demonstrating the same thing, often a lightly revised version of one intervention in a new site. The two can co-occur but are structurally opposite — one churns the what, the other re-proves a fixed what without ever scaling it.

Scope of Application¶

Pilot purgatory lives across the sectors of organizational innovation-adoption where the system that runs pilots is structurally separated from the one that must absorb the scaled capability; these are four sectors of one substrate viewed under different vocabularies, not distinct substrates, and the transition-mechanism diagnosis ports across them intact, while the substrate-neutral conjunction (local success across a missing seam into a stable wrong-level equilibrium) is carried by local_optimum + interface_mismatch + path_dependence, not by this name.

Manufacturing digitization / Industry 4.0 — where the term was coined (McKinsey 2017), industrial firms running dozens of digital-transformation pilots over years without enterprise rollout.
Government technology and digital services — innovation-lab and policy pilots that demonstrate well but never reach agency procurement, statutory authority, or appropriations.
Health-tech adoption — clinical pilots of digital tools or decision support that succeed on-site but never reach the EHR build, a reimbursement code, or clinician workflow.
Education reform — programme pilots in a few schools showing favourable results that never propagate into district- or state-wide adoption, curriculum, or teacher preparation.
Research-into-practice translation — favourably evaluated RCT and implementation-science demonstrations whose adoption pathway is broken.

Clarity¶

Naming pilot purgatory corrects a misdiagnosis that the pattern almost guarantees: the failure looks like insufficient evidence — each pilot ends with a favourable but somehow non-decisive write-up — so the reflex is to run another, better pilot. The label moves the diagnostic question one level up, from "did the pilot succeed?" to "what is the path from a successful pilot to institutional adoption, and does that path exist?" That reframe is load-bearing precisely because the loop is sustained by local success: every cycle does its job, produces a clean result, and that result justifies a repeat rather than triggering escalation. So the concept lets a practitioner stop indicting the pilots — which are working — and locate the fault in the decision infrastructure above them, where a successful demonstration is supposed to convert into a budget commitment, a procurement vehicle, an operating-model change, a named owner.

The reframe is consequential because the two diagnoses prescribe opposite work. "Poor evidence" calls for more rigorous pilots; "missing transition mechanism" calls for stage-gates designed to force a scale-or-kill decision (not permit a third pilot), an adoption owner who holds the operating budget, and a procurement route capable of enterprise rollout — none of which would help if the problem were really evidence quality. The label also sharpens the boundary against neighbors a frustrated funder would blur: it is not sunk-cost persistence (a bias toward continuing because of prior spend — here the path to commitment is structurally absent, even when commitment would be rational), not the valley of death (a financing gap specifically, where pilot purgatory is transition-mechanism-shaped more broadly), and not analysis paralysis (which never decides — pilot purgatory decides favourably, just at the wrong level). The crisp question it leaves on the table is whether the system that runs the pilot is even institutionally connected to the system that must absorb the scaled capability; where the two are separated, no amount of pilot quality closes the gap.

Manages Complexity¶

A funder or program lead staring at a stalled initiative confronts a thicket of candidate explanations for why a string of good pilots never became anything: maybe the evidence was not quite convincing, maybe the technology needs more maturing, maybe the vendor underdelivered, maybe the political winds shifted, maybe the next pilot should be bigger or in a different site. That open-ended diagnostic field, re-litigated every cycle, is the sprawl — and it has a built-in attractor toward "run another, better pilot," because each completed pilot really did succeed within its scope. Pilot purgatory collapses the thicket to a single structural question asked one level above the pilot: does a transition mechanism exist that converts a successful demonstration into institutional adoption — budget ownership, a procurement vehicle capable of enterprise rollout, an operating-model change, a named owner with authority to say yes? The analyst stops grading pilots and instead checks for the presence or absence of that handoff, reading the initiative's fate off whether the path exists rather than off the quality of the latest write-up.

The compression works because it relocates the load-bearing variable. The naive view treats pilot success as the thing to optimize and the distance to scale as an evidence gap to be closed with more rigor; pilot purgatory reveals that the pilot subsystem is already operationally complete — funding vehicle, team, steering committee, evaluation — and that the binding constraint lives entirely in the decision infrastructure above it. Once the question is "is the system that runs the pilot even institutionally connected to the system that must absorb the scaled capability?", the qualitative outcome follows: where those two systems are structurally separated, no increment of pilot quality closes the gap, and the loop will repeat; where the transition handoff exists and is functional, a successful pilot can convert. One presence/absence check on transition infrastructure substitutes for an endless case-by-case audit of evidence sufficiency.

The relocation also installs the branch structure that the misdiagnosis flattens. The decisive fork is evidence problem versus transition-mechanism problem: the first branch calls for better pilots and is the trap; the second calls for entirely different work — stage-gates engineered to force a scale-or-kill decision rather than license a third pilot, an adoption owner who holds the operating budget, a procurement route sized for enterprise rollout. The concept further sharpens boundaries against look-alikes that would otherwise route the analyst to the wrong branch: this is not sunk-cost persistence (the path to commitment is structurally absent, not declined out of bias), not the valley of death (a financing gap specifically, where this is transition-shaped more broadly), and not analysis paralysis (which never decides, whereas pilot purgatory decides favorably, just at the wrong level). What presents as a murky, recurring failure of evidence reduces to one structural locator — present-or-absent transition mechanism — feeding a clean two-way branch that tells the practitioner whether to improve the pilot or to stop running pilots altogether and build the path.

Abstract Reasoning¶

Pilot purgatory licenses reasoning moves that all relocate the diagnosis one level above the pilot — from "did the demonstration succeed?" to "does a mechanism exist that converts success into adoption?" — and exploit the fact that the loop is sustained by local success, not failure.

Diagnostic (infer the structural fault from a signature of repeated success): the counterintuitive central move is to read a string of favourable pilots that never scaled as evidence of a missing transition mechanism rather than weak evidence. The analyst reasons from "every cycle produced a clean, favourable write-up that justified another pilot" to "the pilots are doing their job; the fault sits in the decision infrastructure above them." The signature is precisely the absence of escalation despite local success — a loop that is stable at the pilot state, ending each iteration at the same place with no scale decision triggered. The decisive diagnostic question is institutional connectivity: is the system that runs the pilot even connected to the system that must absorb the scaled capability? Where those two systems are structurally separated — pilot budget in one directorate, operating budget in another; small-grant procurement that cannot scale; no senior owner of the transition decision — the analyst infers that no increment of pilot quality will close the gap, because the gap is not in the evidence. The tell that distinguishes this from genuine evidence weakness is the favourability of the results: pilots that succeed and still do not propagate point upward, not inward.

Interventionist (name the change, predicted to differ sharply from the reflex): the two diagnoses prescribe opposite work, and the concept's force is in predicting that the reflexive fix fails. "Run a better pilot" is predicted to reproduce the loop, because a cleaner result still only justifies another pilot. The predicted-effective interventions target the transition infrastructure: a stage-gate engineered to force a scale-or-kill decision rather than license a third pilot; a named adoption owner who holds the operating budget (so the yes has authority and funding behind it); a procurement vehicle sized for enterprise rollout; and an operating-model change at the receiving organization to absorb the capability. Each is predicted to convert success into adoption only if it bridges the specific separation the diagnostic found — so the interventionist move is to locate which link in the handoff (budget, procurement, governance, ownership) is absent and build that one, with the explicit prediction that pilot-quality improvements will not substitute for it.

Boundary-drawing (separating pilot purgatory from its look-alikes, each of which routes to a different fix): the concept draws crisp lines that prevent misrouting. It is not sunk-cost persistence — the path to commitment is structurally absent, not declined out of bias, so debiasing does nothing. It is not the valley of death — that is a financing gap specifically, whereas pilot purgatory is transition-mechanism-shaped more broadly, and pure financing may be present while the handoff is still missing. It is not analysis paralysis — that never decides, whereas pilot purgatory decides favourably, just at the wrong level. And it is not repeatedly changing direction (pivot thrashing) — it is repeatedly demonstrating the same thing. Drawing these boundaries tells the analyst that the regime in which the pilot-purgatory diagnosis applies is exactly the one where pilots complete and succeed but a structural transition path is missing; outside that regime, a different anti-pattern (and a different cure) is in play.

Predictive / order-of-events: the framing predicts the characteristic non-progression sequence — pilot runs to completion under its own scope, produces a favourable evaluation, the evaluation justifies a repeat rather than triggering escalation, and a fresh pilot is commissioned, often on a slightly revised version of the same intervention, in a new site. Reasoning along that loop, the analyst predicts that the cycle will not self-terminate (there is no escalation pressure built in) and will continue indefinitely until a transition mechanism is installed, so the prediction for any initiative is read off a single structural check: where the runner and the absorber are institutionally separated, expect repetition; where the handoff exists and is functional, expect a successful pilot to convert.

Knowledge Transfer¶

Within organizational change and innovation-adoption practice the pilot-purgatory frame transfers as mechanism, and the transfer between sectors is direct because the load-bearing diagnosis (relocate the fault one level above the pilot, to a present-or-absent transition mechanism) and its toolkit are sector-indifferent. The structural account and its interventions — transition-mechanism audit, budget-gravity analysis, stage-gates with an explicit adoption gate, named adoption owners holding the operating budget, scale-or-kill discipline that refuses a third pilot of the same intervention — carry intact across manufacturing digitization / Industry 4.0 (where the term was coined), government technology and digital services (lab pilots that never reach agency procurement or appropriations), health-tech adoption (clinical pilots that do not reach the EHR build, a reimbursement code, or clinician workflow), education reform (school pilots that never become district- or state-wide), and research-into-practice translation (favourably evaluated demonstrations with a broken adoption pathway). The seed is precise that these are not distinct substrates but four sectors of one substrate — organizational innovation adoption under uncertainty — viewed under different vocabularies, so the breadth is many surface forms of one structure rather than recurrence across structurally distinct ones.

Beyond organizational practice the honest reading is mixed. The occasional off-substrate uses — a long-running research project that never publishes, a startup stuck in perpetual prototype, a perpetual dissertation-stage doctoral student — are analogy (case A): they borrow the "stuck in the demonstration phase" shape, but they are themselves instances of the broader pattern rather than transfers of pilot purgatory's named mechanism, and they carry none of its transition-infrastructure apparatus. What genuinely travels is one abstraction level up, as a conjunction of substrate-neutral parents (case B): stripped of organizational vocabulary, pilot purgatory decomposes into local success at the wrong level (a special case of local_optimum / goal_displacement — the pilot subsystem succeeds at its scoped goal while the higher-level adoption goal goes unaddressed), a missing transition mechanism (a special case of the interface_mismatch / seam pattern between the innovation subsystem and the operations subsystem), and cycle stability at the wrong objective (a stable fixed point, related to equilibrium and path_dependence). So the cross-domain lesson should carry that conjunction of parents, which recurs wherever a subsystem optimizes locally across a missing seam into a stable wrong-level equilibrium. The pilot-purgatory-specific cargo stays home: pilots and proofs-of-concept, scaling decisions, procurement vehicles, governance committees, and budget-ownership structures are organizational-architecture furniture, and the intervention vocabulary (stage-gate design, adoption owners, scale-or-kill, procurement redesign) does not generalize off-substrate without collapsing into those parents. The cleanest disposition is the seed's: keep pilot purgatory as the organizational-management instance, named alongside valley_of_death and sunk_cost_and_irreversible_commitment as distinct innovation-adoption anti-patterns, and cross-linked to local_optimum, interface_mismatch, and path_dependence as substrate-independent parents.

Examples¶

Canonical¶

The defining instance is the Industry 4.0 case McKinsey named when it coined the term around 2017–2018. Surveying manufacturers investing in digital-transformation technologies — sensors, analytics, automation, digital twins — McKinsey found that a large majority (on the order of 70% in its reporting) had run pilots that succeeded at individual plants or lines yet remained stuck, unable to roll the proven capability out across the enterprise. Firms would demonstrate a predictive-maintenance system or a smart-factory line, write up favorable results, and then commission another pilot at another site rather than fund an enterprise deployment. The pilots were not failing; each produced a clean business case within its scope. What was missing was the corporate mechanism — an owner with enterprise budget authority, an operating-model change, a procurement route sized for a company-wide rollout — to convert those local wins into production at scale.

Mapped back: Each plant-level proof-of-concept is the demonstration loop, operationally complete and yielding the successful local evaluation that justifies the next pilot rather than an escalation. The enterprise organization that would have to budget, procure, and operate a company-wide system is the absorber system, structurally disconnected from the pilot teams. The corporate handoff that never existed — enterprise budget ownership and rollout procurement — is the missing transition mechanism.

Applied / In Practice¶

Global digital health calls the same pathology "pilotitis." Across low- and middle-income countries, donor- and NGO-funded mobile-health (mHealth) projects proliferated through the 2010s — SMS reminders, community-health-worker apps, digital diagnostics — each launched as a time-boxed pilot with its own grant, evaluation, and favorable write-up. The World Health Organization and health-systems researchers flagged that these pilots rarely scaled into national health systems: when the grant ended, so did the project, and a fresh donor would fund a near-identical pilot elsewhere. The binding constraint was not evidence that the tools worked but the absence of a route from a successful pilot into the ministry of health's budget, procurement, and operating workflow — no national owner, no integration into the health information system, no recurrent domestic financing.

Mapped back: Each grant-funded mHealth trial is the demonstration loop, and the recurring favorable evaluations that seed the next pilot rather than a national rollout are the cycle stability at the pilot state. The health ministry that would have to fund and operate a national system is the absorber system, severed from the donor-run pilots. The remedy — a national adoption owner holding recurrent budget and a procurement path into the health system — is precisely the counter-mechanism the frame prescribes.

Structural Tensions¶

T1: Pilots working versus pilots as the trap (success is the fuel). The pathology's defining feature is that the loop is sustained by each pilot's local success, not by failure — every cycle produces a clean, favourable write-up that does its job and justifies another pilot. So the healthiest-looking part of the system, the successful demonstration, is precisely the mechanism of the disease. This makes the pathology hard to attack: one cannot simply condemn the pilots, because they genuinely produce evidence and learning, and the success that should be a step toward scale is instead the thing re-justifying the loop. The tension is that local success is at once real value and the engine of stuckness, so the intervention cannot be "run fewer/worse pilots" but must change what a success is allowed to trigger. Diagnostic: Is this pilot generating evidence a scale decision actually needs, or re-proving a known result whose only effect is to re-justify the loop?

T2: Prudent de-risking versus indefinite deferral (the rationale that never expires). Piloting exists for a good reason — de-risk an intervention before committing at scale — and the concept does not say pilots are bad. But the same prudence that justifies a first pilot can be invoked to justify the nth: "one more pilot to be sure" always sounds responsible, so legitimate risk-reduction and pathological repetition are indistinguishable from inside the pilot frame. The rationale that makes piloting wise is the rationalisation that keeps purgatory running, and there is no internal signal that says "enough." The tension is that the virtue of caution has no built-in stopping rule, so the discipline of de-risking degrades into the avoidance of the scale-or-kill decision it was meant to inform. Diagnostic: Is another pilot reducing a specific, decision-blocking uncertainty, or is "be sure first" being used to defer a scale-or-kill decision that the current evidence could already support?

T3: Scale-or-kill discipline versus forced-wrong-decision risk (the cure removes the loop's safety). The prescribed counter-mechanism is a stage-gate that forces a scale-or-kill decision rather than licensing a third pilot. But forcing a binary at a fixed gate has its own failure mode: killing a genuinely promising intervention that needed another iteration, or scaling one prematurely before the evidence warrants. The loop's very stability is a kind of safety — it never commits to a mistake — and scale-or-kill removes that safety in exchange for breaking the stall. The tension is that the mechanism which cures indefinite piloting trades the risk of never deciding for the risk of deciding wrongly under a deadline, and the concept does not itself say when the gate's forcing is warranted. Diagnostic: Does forcing a scale-or-kill decision now improve on continued piloting, or is the loop's caution actually warranted because the scale decision genuinely cannot yet be made responsibly?

T4: Structural relocation versus a real evidence problem (when the pilot won't generalise). The concept's central move is to relocate the fault from evidence quality to a missing transition mechanism, warning that "run a better pilot" is the trap. That relocation is powerful and usually right. But the two problems can co-occur, and the relocation can misfire: sometimes a pilot's success really is non-representative — a hand-picked site, unusually motivated champions, Hawthorne effects — so the demonstrated result genuinely would not survive scaling, and a more representative pilot is the right next step. The tension is that the framing biases toward "it's structural, not evidential," which can dismiss a genuine generalisation failure as a mere missing handoff and wave a fragile result toward a scale it cannot support. Diagnostic: Is the pilot's success genuinely scalable (so the fault is the missing transition), or a non-representative local win that would not survive scaling (so evidence really is the problem)?

T5: Autonomy versus reduction (an innovation-adoption anti-pattern or a local-optimum-across-a-seam). Pilot purgatory is a named organizational anti-pattern with specific cargo — pilots and proofs-of-concept, procurement vehicles, stage-gates, adoption owners, budget-ownership structures — and across its sectors (Industry 4.0, gov-tech, health-tech, education, research translation) it ports as mechanism, though those are four sectors of one substrate, not distinct ones. Stripped of organizational vocabulary it decomposes into a conjunction of parents: local_optimum/goal_displacement (the pilot subsystem succeeds at its scoped goal while the adoption goal goes unmet), interface_mismatch/seam (the missing handoff between the innovation and operations subsystems), and path_dependence/equilibrium (a stable fixed point at the wrong level). Off-substrate uses — a perpetual prototype, an all-but-dissertation student — are analogy. Diagnostic: Resolve toward local_optimum + interface_mismatch + path_dependence when the case is any subsystem optimizing locally across a missing seam into a stable wrong-level equilibrium; toward pilot purgatory for organizational innovation-adoption with its transition-infrastructure apparatus.

Structural–Framed Character¶

Pilot purgatory sits on the framed-leaning side of the structural–framed spectrum: it is not a phenomenon in the world but a named organizational anti-pattern — a pejoratively-framed diagnosis of a human institution's decision infrastructure. On evaluative_weight it leans framed: "purgatory" is a convicting label, and to say an initiative is "stuck in pilot purgatory" is to render a verdict that its decision infrastructure has failed — though the underlying condition (a missing transition mechanism) is describable as a neutral structure, the name carries the charge of a trap to escape. Human_practice_bound is high: the pathology is constituted by the practice of organizational innovation-adoption and dissolves the instant that practice is removed — its whole content is pilots, proofs-of-concept, budget ownership, procurement vehicles, governance committees, and stage-gates, none of which exist outside human organizations; there is no pilot purgatory without an institution that runs pilots and another that must absorb them. Institutional_origin is framed: the concept was coined by McKinsey (around 2017) as an Industry 4.0 diagnosis and lives in management-consulting and innovation-management discourse — an artifact of a practitioner tradition, not a distinction nature draws. On vocab_travels it fails for its named cargo: pilots, scale decisions, procurement redesign, adoption owners, and scale-or-kill stage-gates are organizational-architecture furniture that does not survive extraction. Import_vs_recognize is bimodal: across its sectors (manufacturing digitization, gov-tech, health-tech, education reform, research translation — four sectors of one substrate) the mechanism is recognized intact, while beyond it the off-substrate uses (a perpetual prototype, an all-but-dissertation student) are analogy, borrowing the "stuck in demonstration" shape without the transition-infrastructure apparatus.

Pilot purgatory is a case where the portable content is not a single skeleton but a conjunction of substrate-neutral parents that the concept composes: a subsystem succeeds at its scoped local goal while the higher adoption goal goes unmet (local_optimum/goal_displacement), across a missing handoff between the innovation subsystem and the operations subsystem (interface_mismatch/seam), settling into a stable wrong-level equilibrium that will not self-terminate (path_dependence/equilibrium). That conjunction is what pilot purgatory instantiates from its parents, not what makes "pilot purgatory" itself travel: the cross-domain reach — any subsystem optimizing locally across a missing seam into a stable wrong-level equilibrium — belongs to local_optimum + interface_mismatch + path_dependence, while pilot purgatory's distinctive cargo (the pilots, the procurement vehicles, the scale-or-kill governance) stays home in organizational innovation-adoption. Its character: a practice-constituted, discourse-originated, pejoratively-named organizational anti-pattern, structural only in the local-success-across-a-missing-seam-into-a-stable-wrong-level-equilibrium conjunction it borrows from local_optimum/interface_mismatch/path_dependence and frames as an innovation-management trap.

Structural Core vs. Domain Accent¶

This section decides why pilot purgatory is a domain-specific abstraction and not a prime, and it carries the case for its domain-specificity in one place. Its portable content is a conjunction of parents, so the skeletal core must be named as several primes at once.

What is skeletal (could lift toward cross-domain primes). Strip the organization and a thin relational structure survives, composed of several primes: a subsystem succeeds at its scoped local goal while the higher-level goal goes unmet, across a missing handoff between the demonstrating subsystem and the absorbing one, settling into a stable wrong-level equilibrium that will not self-terminate. The portable pieces are abstract — local success at the wrong level (local_optimum/goal_displacement), a missing seam between two subsystems (interface_mismatch/seam), and a self-perpetuating fixed point (path_dependence/equilibrium). That conjunction is genuinely substrate-portable, recurring wherever a subsystem optimizes locally across a missing seam into a stable wrong-level equilibrium, which is exactly why the entry instantiates those three parents. But it is the core the entry shares, not what makes pilot purgatory distinctive.

What is domain-bound. Almost everything that makes the concept pilot purgatory in particular is organizational-innovation furniture, and none of it survives extraction. The demonstrating subsystem is specifically a pilot / proof-of-concept (with a funding vehicle, team, steering committee, and evaluation write-up); the absorbing subsystem is an organization that must budget, procure, operate, and integrate the scaled capability; the missing handoff is a transition mechanism — budget ownership, a procurement vehicle, governance, a named adoption owner; the failure signature is a favourable evaluation that justifies another pilot rather than triggering a scale decision; and the prescribed counter-mechanism is a scale-or-kill stage-gate and enterprise-sized procurement. The decisive test: strip the pilots, the procurement vehicles, the scale-or-kill governance, and the budget-ownership structures — keeping only "local success across a missing seam into a stable wrong-level equilibrium" — and it is no longer pilot purgatory but the bare conjunction of its parents, because the innovation-adoption apparatus that gives the diagnosis its content and its cure has been stripped away. Off-substrate uses (a perpetual prototype, an all-but-dissertation student) borrow the "stuck in demonstration" shape but carry none of that apparatus. The concept is constituted by the organizational-innovation context the prime bar asks it to shed.

Why this does not clear the prime bar. A prime is a relational structure whose vocabulary travels and whose cross-domain transfer is recognition of the same mechanism, not analogy. Pilot purgatory's transfer is bimodal. Within organizational innovation-adoption it travels intact as mechanism — manufacturing digitization / Industry 4.0, government technology and digital services, health-tech adoption, education reform, and research-into-practice translation are four sectors of one substrate viewed under different vocabularies, so the transition-mechanism diagnosis, the evidence-versus-transition fork, and the counter-mechanism toolkit re-apply without translation. Beyond organizations it travels only by analogy: a perpetual prototype or a perpetual dissertation borrows the demonstration-stuck shape but is itself an instance of the broader pattern, not a transfer of pilot purgatory's named mechanism. And when the bare structural lesson is needed cross-domain — a subsystem optimizing locally across a missing seam into a stable wrong-level equilibrium — it is already carried, in more general form, by the local_optimum + interface_mismatch + path_dependence conjunction the entry composes. The cross-domain reach belongs to those parents; "pilot purgatory," as named — the pilots, the procurement vehicles, the scale-or-kill governance, the adoption owners — carries innovation-adoption baggage that does not and should not travel. (The cleanest disposition keeps pilot purgatory as the organizational-management instance, named alongside valley_of_death and sunk_cost_and_irreversible_commitment as distinct anti-patterns and cross-linked to those three parents.)

Relationships to Other Abstractions¶

Current abstraction Pilot Purgatory Domain-specific

Parents (1) — more general patterns this builds on

Pilot Purgatory is a kind of Interface Mismatch Prime

Pilot Purgatory is the organizational interface mismatch in which a demonstration system's outputs do not satisfy the absorber system's adoption contract.

Hierarchy path (1) — routes to 1 parentless root

Pilot Purgatory → Interface Mismatch → Interface → Boundary

Not to Be Confused With¶

Valley of death. A financing gap specifically — the funding chasm between R&D and commercialization where a promising technology dies for lack of capital to cross from prototype to market. Pilot purgatory is transition-mechanism-shaped more broadly: funding may be fully present while the missing link is procurement, governance, operating-model integration, or budget ownership. Financing is at most one of several handoffs that can be absent. Tell: is the sole missing link money to bridge R&D-to-market (valley of death), or is capital available while a non-financial handoff to adoption is missing (pilot purgatory)?
Sunk-cost persistence / irreversible commitment. A bias toward continuing an initiative because of prior spend. Pilot purgatory is the structural absence of the path to commitment even when commitment would be rational — no one is irrationally clinging to a doomed program; the route to scale simply does not exist. Debiasing does nothing here. Tell: is the initiative being continued against the evidence out of attachment to past investment (sunk cost), or is the pathway to scale structurally missing regardless of anyone's attachment (pilot purgatory)?
Analysis paralysis. A decision pathology that never decides — endless deliberation with no verdict. Pilot purgatory decides favourably, just at the wrong level: each cycle reaches a clean positive verdict on the pilot and acts on it by commissioning another pilot. The failure is not indecision but a decision that loops at the demonstration level instead of escalating. Tell: is the system stuck before any verdict (analysis paralysis), or repeatedly reaching a positive verdict that triggers a repeat rather than escalation (pilot purgatory)?
Pivot thrashing. The structural opposite (and a sibling entry): repeatedly changing direction, churning the "what." Pilot purgatory repeatedly demonstrates the same thing, often a lightly revised version of one intervention in a new site — re-proving a fixed what without ever scaling it. The two can co-occur but invert each other. Tell: is the initiative changing its target each cycle (pivot thrashing) or re-proving a stable target without ever committing to scale (pilot purgatory)?
Innovation theater / performative piloting. Pilots run mainly to signal innovativeness, with no genuine intent to scale — the demonstration is the goal, not a step toward adoption. Pilot purgatory wants to scale and would benefit from doing so; it is blocked by a missing transition mechanism, not by an absent intent. Tell: is scaling never actually desired, the pilot being its own performative end (innovation theater), or is scaling genuinely wanted but structurally unreachable (pilot purgatory)?
The local_optimum + interface_mismatch + path_dependence conjunction. The substrate-neutral parents pilot purgatory composes — a subsystem succeeding at its scoped local goal, across a missing seam to the absorbing subsystem, settling into a stable wrong-level equilibrium. Off-substrate look-alikes (a perpetual prototype, an all-but-dissertation student) are analogy: they borrow the "stuck in demonstration" shape but carry none of the transition-infrastructure apparatus. Tell: the conjunction carries the local-success-across-a-missing-seam lesson anywhere; "pilot purgatory" is reserved for organizational innovation-adoption with its pilots, procurement, and scale-or-kill governance.

Neighborhood in Abstraction Space¶

Pilot Purgatory sits in a sparse region of the domain-specific corpus (93^rd percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Incident Command & Operational Tempo (10 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-07-12